Comments on the paper « Basin inversion: Reactivated rift structures in the Ligurian Sea revealed by OBS » by M. Thorwart et al., submitted to Solid Earth.

General comments

The paper by Thowart et al. presents the seismotectonic interpretation of the data acquired in the Ligurian basin during 8 months in 2017-2018 by the 24 OBS of the Alparray experiment. This is the first data set of such importance acquired in this basin, with atypical oceanic crust or very thinned continental crust, located between the front of the southern subalpine chains and the Corso-Sardinia continental block. This area is known since a long time as an active intraplate setting with a very low deformation rate. 

The experiment allows to detect 39 microearthquakes. This work focus on two clusters in the center of the basin, the first one with 13 events and the second with 3 events but only 4 focal mechanisms could be determined among the 13 events of the cluster 1. These 4 focal mechanisms, consistent with each other, are interpreted by the authors as evidence of the Ligurian basin inversion.

There is little new data but in this marine and low seismicity context any new data is welcome to be discussed and should be considered positively. The active inversion of the Ligurian basin has already been evidenced based on other structural and seismic datasets (Larroque et al., 2011; Sage et al., 2011; Larroque et al., 2016) and from this perspective this work confirms what is proposed. However, it seems to me that there are several important problems of substance and form in this article which require a serious reworking of the presentation of the data and their interpretation in the context of the work already carried out in this area. I recommend major revision.

I make few remarks and suggestions in the following.

- On the substance, the major problem concerns the exploitation of the seismic signal from very low magnitude earthquakes (1.5 to 2.5) recorded at stations located more than 150 km away (Fig1) to build the focal mechanisms. This essential part of the work must be strengthened to be credible. Firstly, as these are new mechanisms it is necessary to provide for each of them a clear diagram with the polarities and nodal planes (currently, only the diagram of the strongest earthquake is shown : c, fig4). Secondly, taking into account the large distances with some of the stations and the smallness of the magnitudes, the seismograms must be shown in order to attest the quality of the polarities read on this signal.

- The context of the deformation in the Ligurian basin must be presented in its entirety. Particularly, the high rate of seismicity on the northern margin in relation to the center of the basin and its southern margin must be emphasized as it is an essential point in the discussion of the inversion process. From this point of view, mention of the work of Béthoux et al (2008) is essential. Also from a structural point of view, it should be mentioned that active north dipping reverse faults have been identified on the northern margin (Larroque et al., 2011; Sage et al., 2011). These faults allow the accommodation of most of the basin inversion since 5 Ma, as evidenced by the cumulative deformation which shows a margin uplift of more than 1000 m with respect to the basin. These 2 points are important because they show that the basin inversion started at least 5 Ma ago in the northern part while the absence of cumulated deformation and low seismicity in the central and southern part (this paper and Larroque et al., 2016) attest to a weaker and/or much more recent deformation.

- The input of new data is really low for such a paper in a major scientific journal. May be the authors could try to get more informations from the continuous seismic recordings of the OBS by using template matching method ? Even if the signals are not usable to determine focal mechanics, it would be interesting to know if a larger number of low magnitude events can be detected.

- You need to take better account of existing work when it relates to your interpretations. For instance, the proposal by Dannowski et al (2020) on the nature of the crust in the Ligurian Basin is very interesting but at the moment it is not a consensual result. So highlight other interprétations such as Contrucci et al. (2001), Rollet et al. (2002), Gailler et al. (2009).

- The use of the results of the Pérez-Guissinyé and Reston model does not seem to me very adapted to the case of the Ligurian rifting. This model has been proposed by its authors to describe a possible evolution of Atlantic-type rifted margins in the case of cratonic and old orogen models. It is therefore difficult to consider that this model is generalizable to all types of non-volcanic rifting. The Ligurian basin is a back-arc basin, the crust was strongly affected by the alpine orogeny. The initial rheological conditions are therefore strongly different from what is considered in the Pérez-Guissinyé and Reston model. This comparison should therefore be discussed and justified.

Specific comments

- Revise the title because most of the inversion is active on the northern margin that is not discussed. For the moment, only a recent and low compression is carried out in the center of the basin.

line 27 : usually, moderate activity is less than magnitude 6 and strong activity start with Mw 6.5 thus the 1887 Ligurian eq (Mw > 6.5 Larroque et al., 2012, Manchuel et al., 2017) attests that the sismicity on the Ligurian margin is mainly moderate but with possible strong earthquakes (this is of concern for hazard assement).

Line 51 : the geodynamic setting WAS controlled by the Africa/Euraisa convergence, now it is not so clear (see Nocquet and Calais, 2004, Serpelloni et al., 2007, ….).

Line 54 : « Lamotte » is Frizon de Lamotte.

Line 70 : specify the rotation pole near Genoa and give the range of the counter-clockwise rotation from 23° (Speranza) to 45° (Gattacceca et al., 2007). 

Line 159 : can you explain « faulty recording » ?

Line 34 and 190-196 : I disagree with this presentation of the spatial geodetic data. Nocquet and Calais (2004) showed that the convergence of Africa relative to stable Europe is 40% less than the prediction of the Nuvel-1 model (De Metz et al., 1994). Serpelloni et al. (2007) and Nocquet (2012) argue that 90% of the convergence is accomodated along the Maghrebides chain and Algeria margin but ~10% of the motion could be accommodated northward with a possible NW motion of the Corsica–Sardinia bloc of 0.5 mm/yr maximum. Masson et al. (2019) confirmed a NW motion of Corsica in the range of 0.4 mm/yr.

Line 194 : In my opinion reference to these long-term plate models (van Hinsbergen, Le Breton) are not relevant to discuss current movements because they have no resolution for the present-day.

Line 201 : the dip of the ~vertical plane (d, Fig5) is NW not SE ? In any case, in order to discuss the dip, one must have information on the uncertainties of epicenter location and depth because the difference are very tenuous.

Line 215-219 : you should mention the more recent and more precise work carried out on the rotation of the Corsica-Sardinia block by Gattacceca et al (2007). The rotation reconstructed from the paleomagnetic analyses of lavas is 45° counter-clockwise which leads to a much greater extension in the basin than from the 23° proposed by Speranza et al. (2002). At least both values should be mentioned.

Line 248-253 : the discussion about heat flow and the role of the sedimentary cover should take into account the results of Béthoux et al. (2008). 

Figure 2 : this figure is to be improved by showing the Ligurian thrust to the north, the normal faults of the Tyrrhenian Sea and the thrust to the north of Algeria….

Caption figure 5 : (d) « crystalline basement » = top of crystalline crust ; number on (e) = Vp ?

Technical corrections

Check the references, e.g. Maggi et al. ???

Gattacceca, J., Deino, A., Rizzo, R, Jones, D.S., Henry, B., Beaudoin, B., Valeboin, F., 2007. Miocene rotation of Sardinia : new paleomagnetic and geochronological constraints and geodynamic implications. Earth Planet. Sci. Lett., 258, 359–377.

Larroque, C., Mercier de Lépinay, B., Migeon, S., 2011. Morphotectonic and fault–earthquake relationships along the northern Ligurian margin (western Mediterranean) based on high resolution multibeam bathymetry and multichannel seismicreflection profiles. Mar. Geophys. Res. 32 (1–2), 163–179, http://dx.doi.org/10.1007/s11001-010-9108-7.

Manchuel, K., Traversa, P., Baumont, D., Cara, M., Nayman, E. and Durouchoux, C., 2017. The French seismic CATalogue (FCAT-17). Bull. Earthquake Eng., 8, 16. 2227–2251, doi: 10.1007/s10518-017-0236-1.

Masson, C., Mazzotti, S., Vernant, P., 2019. Precision of continuous GPS velocities 20 from statistical analysis of synthetic time series. Solid Earth, 10, 329–342, https://doi.org/10.5194/se-10-329-2019.

Sage, F., Beslier, M.O., Thinon, I., Larroque, C., Dessa, J.X., Migeon, S., Angelier, J., Guennoc, P., Schreiber, D., Michaud, F., Stéphan, J.F., Sonnette, L., 2011. Structure and evolution of a passive margin in a compressive environment : example of the south-western Alps-Ligurian basin junction during the Cenozoic. Mar. Pet. Geol., 28, 1263–1282, doi:10.1016/j.marpetgeo.2011.03.012.

Serpelloni, E., Vannucci, G., Pondrelli, S., Argnani, A., Casula, G., Anzidei, M., Baldi, P., Gasperini, P., 2007. Kinematics of the Western Africa-Eurasia plate boundary from focal mechanisms and GPS data. Geophys. J. Int., 169(3), 1180-1200, http://dx.doi.org/10.1111/j.1365246X.2007.03367.x.

This manuscript is a short paper presenting seismic activity in the Ligurian Sea, detected by the recently deployed AlpArray seismic network (offshore component LOBSTER), and therefore fits very well with this special edition of Solid Earth. The data and results are very interesting as it shows clusters of compressive earthquakes in the center of the Ligurian Basin, suggesting inversion of this part of the basin, and at intriguing high depths (10-16 km; lower crust - upper mantle). However, the authors need to better present/discuss the uncertainties of the method applied and their results, and improve the discussion of their results in terms of rift-related structures and strength of the lithosphere (see main comments below). Therefore, I suggest major revision.

Main comments:

- Section 3 - Data, methods & results: Section 3 lacks information on the uncertainties of the method applied and the results, especially regarding the depth of earthquakes and the orientation of the fault planes, which are essential to assess/support the interpretation in terms of rift-related structures and rheology that follows in section 4. Table 1 provides uncertainty range on the depth of each event but it is not explained how this was estimated?

- Section 4.2 – Orientation pre-existing rift-related faults: the title of section 4.2 should be changed. This studies provides insight on the present-day seismic activity of the basin but not on the rifting history. The authors discuss here the orientation of the faults obtained from the focal mechanisms of the earthquakes and that they interpret to be inverted, pre-existing rift-related faults. I agree, a change of strike from c. SW-NE to more SSW-NNE may reflect (not ‘mimic’) the rotation of normal faults formed during rifting during the CCW rotation of Corsica-Sardinia towards the SE relative to Europe. But this paragraph needs rephrasing and more information should be given, e.g. the exact strike (and dip?) of the fault planes obtained from the focal mechanisms. The authors should also quote papers concerning the evolution of continental rifting that shows younging of rifting-related structures towards the center of the basin (e.g. Brune et al. 2014, Nature communications) to support their interpretation. Also, the rotation of Corsica-Sardinia, was not only between 21-16 Ma but started with continental rifting at about 35-30 Ma, as indicated by the age of syn-rift sediments (e.g. Séranne 1999, J. Geol. Soc. London; Jolivet et al. 2015, Tectonics; note the total amount of rotation between 35-0 Ma is estimated to a minimum of c. 53°, Le Breton et al. 2017). The phase between 21-16 Ma is often interpreted as the phase of oceanic spreading in the Liguro-Provencal Basin (e.g. Speranza et al. 2002; Le Breton et al. 2017), as suggested by the age of post-rift sediments along the margin of Gulf of Lion (e.g. Séranne 1999; Jolivet et al. 2015) and Sardinia (e.g. Sowerbutts and Underhill, 1998, J. Geol. Soc. London). 

- Section 4.3 – Rheology of the lithosphere: This section is important as the depth and location of the earthquakes in the upper mantle, in the middle of the basin is indeed quite intriguing. First, the author should better present the uncertainty on the nature of the crust and depth of crystalline basement (CB on Figure 5) in this part of the basin. Then, they discuss the seismicity in terms of strength of the lithosphere. However, Handy & Brun (2004, EPSL) discussed that seismicity is an ambiguous indicator of strength and proposed that earthquakes are more reasonably interpreted as a manifestation of transient mechanical instability within shear zones and may be used to locate active weak zones within the continental lithosphere. This should be mentioned and discussed here, as shear zones commonly form through lower crust-upper mantle during continental rifting (see for example Naliboff et al. 2017, Nature communications). Finally, when discussing the strength of the lithosphere, temperature is indeed an important parameter but the discussion here is not clear. Why does stretching of the lithosphere would cause “lower” temperature at the crust-mantle boundary (l. 229)? Stretching involves thinning of the lithosphere thus increasing of the geotherm, which is reflected by higher heat flow, as mentioned later in the text (l.248-253). This is contradictory and should be clarified. 

Specific comments:

Abstract (lines 20-22) and Conclusions (lines 259-263) need to be improved/rephrased (see main comments above). For example, I would not say ‘away from the abandoned rift’ but away from the center of the rift basin; ‘mimic’ -> ‘may reflect’ 

Line 35: the entire basin might be under compressive stress but the inversion is mostly observed along the margins of the Tyrrhenian Basin 

Line 45: 8 months – precise 2017-2018

Line 52: rollback of the Apennines, Calabrian and Gibraltar subduction zones

Lines 53-54: south-eastward migrating (or retreating) Apennines-Calabrian arc (Frizon de Lamotte et al.)

Line 56: Alboran Basin: c. 25 – 8 Ma (see for example Comas et al. 1999, Proceedings of the Oceanic Drilling Program, Scientific Result, Vol. 161)

Lines 62-63: I would call it Liguro-Provencal Basin (SW part), not Balearic Sea. Gailler et al. 2009; Afilhado et al. 2015; Moulin et al. 2015 are geophysical studies between Gulf of Lion – Sardinia, which is the southwest part of the Liguro-Provencal Basin. And they mentioned “atypical” oceanic crust, it should be mentioned. 

Line 70: Gattacceca et al. (2007, EPSL) should be also mentioned here.

Lines 73-75: please add time constraints here, since when the opening of the Tyrrhenian Sea ceased? 

Lines 75-76: today

Lines 180-181: The uncertainty on the depth of the crystalline basement (CB on Figure 5), especially for this part of the profile where the distinction between sediments and thinned continental crust cannot be done (Dannowski et al. 2020), should be mentioned and discussed (here and/or in section 4).

Line 191: How does the counter-clockwise rotation of Adria would generate regional compression in the Ligurian Basin? Larroque et al. (2016) discuss the southward propagation of the deformation from the Alps-Ligurian basin junction to the southern margin of the basin, for me this goes in (1) Africa-Europe convergence.

Lines 193-194: Le Breton et al. (2017) and van Hinsbergen et al. (2020) are not really relevant here. Our plate reconstructions do not indicate where the Europe-Adria convergence is accommodated today, but provide information on the long-term plate motion and kinematics. They would be more relevant in geological setting, when presenting the geodynamic evolution of the area over the last 35 Ma.  

Line 195-196: what about the inversion of the northern margin of the Ligurian Basin? (as mentioned in section 2, see also Billi et al. 2011, Bulletin de la Société Géologique de France 182).

Sentences lines 182-183, 200-201 and 203-204: This is not very clear, do the authors suggest that these earthquakes occur along one fault plane (as projected in their profile AB, l. 200-201) or along different fault planes (as mentioned l. 182-183; 203-204)? And why (based on what arguments/observations)? The rupture lengths/areas must be small, as indicated by the low magnitude of these earthquakes, and may explain why the post-rift sediments are not affected. But how does it tell more information on the location of theses earthquakes along one or more fault plane(s)? 

Line 200-201: ‘push’ direction -> slip direction (?)

Line 204: ‘can be taken up by these remaining’ -> could reactivate pre-existing ; ‘rifting structures’ -> rift-related structures (check throughout the text); ‘enabling’ -> suggesting ongoing

l. 216: ‘turned’ -> inverted

l. 224: water in the formation of ?

l. 234: interpreted to reflect inversion along pre-existing normal faults

Figures

Figure 2: I would remove Balearic Sea and keep only Liguro-Provencal Basin (the authors could add the North Balearic Transform Zone, e.g. van Hinsbergen et al. 2014, Tectonics, as a delimitation between the Liguro-Provencal Basin and the Algerian Basin). I suggest also to add on this map the (inferred) location of oceanic crust (or atypical crust) vs exhumed mantle, and zones of basin inversion described in previous work and mentioned in the text.

Figure 5: It would be interesting to plot the clusters of earthquakes directly on the seismic velocity model (e). At about 30 km distance, the velocity contour 5.5 km/s deepens, does the location of cluster C1 coincides with this change in velocity? 

References

Bethoux 1992: Quaternaire (not quate)

Maggi et al. 2000: Journal (Geology) is missing